Market data spend and usage management software provider TRG Screen has launched an artificial intelligence-powered capability to help financial institutions better manage spiralling data costs.
The conversational AI interface sits on top of TRG Screen’s established Xmon platform, allowing users to interact with their own programme data using natural language. Instead of digging through technical reports, users can ask the system direct questions about cost optimisation opportunities and receive actionable recommendations in seconds.“It’s a bit like having an Xmon support analyst available to you 24/7,” Amjad Zoghbi, Head of AI at TRG Screen, told Data Management Insight. “The real power is when you’re conversing with it to achieve a certain outcome. From end-to-end, a single conversation can save a client between 50 minutes and an hour of manual analysis, time they can redirect towards strategic vendor negotiations of proactive cost management.”
The tool operates on three core pillars: cost transparency, optimisation and governance. By analysing massive amounts of pre-calculated usage data, the AI identifies duplicated requests, such as multiple systems requesting the same static data, and inefficient field-level usage that would take a human analyst hours to uncover. These insights translate directly into reduced vendor fees and more efficient data consumption patterns, TRG Screen said.
Under the Bonnet
Xmon AI Assist was developed using a Retrieval-Augmented Generation (RAG) framework. This architecture combines the reasoning power of enterprise-grade Large Language Models (LLMs) — from providers like Anthropic and Microsoft — with each client’s specific Xmon analytics.
The system does not use sensitive client data to train its models. The AI acts as an orchestrator: it understands the user’s intent, retrieves relevant data via a secure REST API and uses that context to provide an answer only for the duration of that session.
“The AI provides the answer but also points you to the actual facts and the source report,” Christophe Plouvier, Xmon Product Director at TRG Screen told Data Management Insight. “The client can go into the applications to justify the answer and see the analytics used.”
Rising Costs
The release comes as financial institutions face climbing data costs amid a market shift from one of a few major players with relatively straightforward flat-fee subscriptions to a complex multi-vendor ecosystem where firms juggle overlapping subscriptions from the likes of Bloomberg and LSEG.
Vendors have moved towards highly granular, volume-based pricing with multiple dimensions. This complexity makes it nearly impossible for humans to track every penny in real-time.
“It is not getting simpler from a compliance standpoint, and it’s not getting cheaper,” said Zoghbi. “Without a system that can reliably and automatically provide suggestions and keep you in control, you cannot remain competitive.”
History of Innovation
TRG Screen has introduced increasingly advanced solutions that have moved market data teams from passive to proactive. Through a combination of strategic acquisitions and proprietary development, the company has created a suite of products for managing usage, spending and vendor compliance. Its most recent acquisition, of Crizit, further expands the company’s offering in data caching and other areas crucial to effective market data programme management.Xmon AI Assist marks the next step in TRG Screen’s AI ambitions, following the successful rollout of similar technology within PEAR, the company’s global exchange policy knowledge base. The company is now moving towards an “intelligent” era of democratising access to insights that were previously only available to senior data specialists.
Future Evolution
TRG Screen plans to move beyond the “question and answer” format in which the AI won’t just suggest a cost-saving rule – it will ask the user for permission to execute it.
“Our AI strategy is driven by client impact and real-world outcomes,” said Zoghbi. “We design every capability to remove friction and bring intelligence directly into the workflows that matter most. This is the very beginning of a big change in the way people interact with our systems.”
Subscribe to our newsletter
